The phosphole ring system is described by structure A. This structure is distinct from the class of compounds B which contain benzo rings fused to the phosphole core. ##STR1## Class I has a much different electronic structure and therefore has much different chemistry than compounds of class II. In class I, the P atom is part of the delocalized, partially aromatic ring system. In class II, the aromaticity is confined to the benzo rings, with no delocalization around the P atom. Class I will participate in Diels-Alder chemistry (especially when complexed to a metal) (Bhaduri et al., Organometallics 1992, 11, pp. 4069-4076), whereas compounds of class II will not (Quin, Compr. Heterocycl. Chem. II Bird, Clive W (Ed), 1996, Vol. 2, pp. 757-856).
Very few compounds have been reported that contain two phosphole rings connected via a bridge (A) between the phosphorus atoms (structure C) (A=bridging hydrocarbon, hydrocarbon/heteroatom(s), or organometallic group). ##STR2##
One explanation for the paucity of compounds of type C-G is the lack of synthetic procedures broad enough in scope to prepare the phosphole ring system. ##STR3##
Examples reported in the literature include the compounds 1 (Braye et al., Tetrahedron 1971, pp. 5523-37), 2 (A=--CH.sub.2 --, --CH.sub.2 CH.sub.2 --, and --CH.dbd.CH--CH.dbd.CH--) (Charrier et al., Organometallics 1987, 6 pp. 586 91), and 3 (Gradoz et al., J. Chem. Soc. Dalton Trans. 1992, pp. 3047-3051). ##STR4##
Compounds containing a single phosphole ring (I) were made using the Fagan-Nugent heterocycle synthesis (Fagan et al., J. Am. Chem. Soc. 1994, 116, pp. 1880-1889; Fagan et al., J. Am. Chem. Soc. 1988, 110, pp. 2310-2312). This synthesis involves preparing the zirconium reagents by coupling of acetylenes followed by transfer of the metallacycle from zirconium to phosphorus. In all cases, the substituent on the phosphorus was an aromatic group such as phenyl.
These types of compounds (containing a single phosphole ring) have found limited utility as ligands for transition metals for use in catalysis, and have been shown to have different chemistries than their phosphine analogs (Neibecker et al., New J. Chem. 1991, pp. 279-81; Neibecker et al., J. Mol. Catal. 1989, 57 pp. 153-163; Neibecker et al., J. Mol. Catal. 1989, 219-227; Vac et al., Inorg. Chem. 1989 28, pp. 3831-3836; Hjortkjaer et al., J. Mol. Catal. 1989 50, 203-210).
Transition metal complexes have been made using structures of class VII, shown below, where the rings are linked at the position alpha to phosphorus. Attempts to use these ligands to make Pd acetonitrile complexes analogous to those in the instant invention failed (Guoygou et al., Organometallics 1997, 16, 1008-1015). ##STR5##
Copolymers of carbon monoxide and olefins, such as ethylene, can be made by free radical initiated copolymerization (Brubaker, J. Am. Chem. Soc., 1952, 74, 1509) or gamma-ray induced copolymerization (Steinberg, Polym. Eng. Sci., 1977, 17, 335). The copolymers produced were random copolymers and their melting points were low. In 1951, Reppe discovered a nickel-catalyzed ethylene carbon monoxide copolymerization system that gave alternating copolymers (U.S. Pat. No. 2,577,208 (1951)). However, the molecular weights of these polymers were also low.
In 1984, U.S. Pat. Nos. 4,818,810 and 4,835,250 disclosed the production of alternating olefin carbon monoxide copolymers based on Pd(II), Ni(II) and Co(II) complexes bearing bidentate ligands of the formula R.sub.1 R.sub.2 E--A--E--R.sub.3 R.sub.4, wherein R.sub.1, R.sub.2, R.sub.3, R.sub.4, and A are organic groups and E is phosphorus, arsenic, or antimony. When E is phosphorus and R.sub.1-4 are aryl groups, the corresponding diphosphine palladium complexes are active in copolymerizing ethylene and carbon monoxide to produce copolymers of molecular weight up to 30,000 (MW.sub.n) (Drent et al., Chem. Rev., 1996, 96, 663). No compounds were claimed or disclosed in which R.sub.1 and R.sub.2, and R.sub.3 and R.sub.4 together formed a ring. Applicants have recently found that the diphosphole coordinated palladium catalysts catalyze olefin/carbon monoxide (CO) copolymerization. When the P atom is part of a ring system, the electronic environment and therefore expected chemistries are different than simple, non-ring phosphine disclosed in the patents described above.
Radical polymerization is an important commercial process for making a variety of polymers of vinyl monomers, such as acrylics and styrenics. While this process makes large amounts of polymers, the difficulty in accurately controlling the polymer structures (such as molecular weight, molecular weight distribution, and architecture, etc.) has significantly limited its further applications.
Living polymerization usually offers much better control on polymer structures and architectures. While living polymerization systems for anionic, cationic, and group transfer mechanisms were developed some years ago, a true living radical polymerization system is still an elusive goal (because of the high reactivity of free radicals) and only very recently has pseudo-living radical polymerization been achieved. One pseudo-living radical polymerization method is "atom transfer radical polymerization" (ATRP). In this process a transition metal compound, usually in a lower valent state, is contacted with a compound which is capable of transferring an atom to the metal complex, thereby oxidizing the metal to a higher valent state and forming a radical which can initiate polymerization. However, the atom that was transferred to the metal complex may be reversibly transferred back to the growing polymer chain at any time. In this way, the propagation step is regulated by this reversible atom transfer equilibrium and statistically all polymer chains grow at the same rate. The results a pseudo-living radical polymerization in which the molecular weight may be closely controlled and the molecular weight distribution is narrow.
Such ATRPs are described in many publications (Kato et al., Macromolecules 1995, 28, 1721; Wang et al., Macromolecules 1995, 28, 7572; Wang et al., Macromolecules 1995, 28, 7901; Granel et al., Macromolecules 1996, 29, 8576; Matyjaszewski et al., PCT WO 96/30421). The transition metal complexes used include complexes of Cu(I), Ru(II), Ni(II), Fe(II), and Rh(II). The complexes are formed by coordinating the metal ions with certain ligands such as nitrogen or phosphine containing ligands. For Ru(II) and Fe(II), mono-phosphine P(C.sub.6 H.sub.5).sub.3 was used as the ligand. However, for Cu(I), all the ligands used are nitrogen-based such as bipyridine or substituted bipyridine. No phosphine-based ligand has been shown to be an effective ligand for Cu(I) in ATRP.
It has been found that novel types of ligands containing phosphole and other P ring systems can chelate Cu(I) to form active catalysts for ATRP.